ST setups are soooo brittle

I’m a pretty casual user and love ST for its integrations and programmability. But… It seems like not a week goes by without something deciding not to work.

  • Two ge-links don’t connect anymore
  • My lifx bulbs start turning purple instead of warm white
  • My wife’s iPhone presence sensor disappears
  • My custom smartapp to light an indicator lamp when the mail arrives stops working
  • A dead battery in a door sensor is reported as a door that just never opens

I can fix all of these things individually, of course… I’m just wondering, generally, does stuff break for everyone this often? It’s a little frustrating for me, and I don’t mind tinkering. I can’t imagine what a more general consumer would think.

yes. It’s a way of life.

I try to minimize my maintenance to about once a week, using SimpleDeviceViewer to review the zombies and AWOLs

For me, the only thing that’s creating an issue is an ST Multi Sensor that I use on a door. I am replacing it out this week. Otherwise I haven’t had to touch my system in months. I’ve had more issues with Google Home than ST.

1 Like

Because I have to pay other people to do any maintenance at all (I am quadriparetic) I pay pretty close attention to this stuff.

In particular, I look for an MFOP (maintenance free operating period) of at least six months and preferably a year.

I get that from the Phillips hue bridge, Logitech Harmony, Amazon echo, Lutron Caseta, HomeKit and some other systems.

However, since October 2015 I have yet to get more than nine days from SmartThings without some kind of maintenance being required, even if it’s just popping the batteries on the sensor, or opening the app and saving something again.

Many of the issues are small with minimal work around, and some get fixed within A few hours or a few days. Some issues persist for months. But I’m sure there are many people who hardly even consider them problems, it’s just that the system is a bit fiddly and needs continual amounts of small maintenance.

For me, though, there’s more to it than that since I have limited use of my hand so at the present time I would say this is a high maintenance system although each individual maintenance effort is fairly small.

I think it depends on what you’re doing. My ST setup has been robust and reliable for ages now, with the only recent hiccup being last weekend when the local power grid had a slew of downtime instances. And then I only had to remove power from both the hub and my Schlage lock, and restart, and they came up correctly.

Then again, I’m not using ST as the be-all of my smart home. It is one component of that home. A central one to be sure, but given some of what I read here I’m not straining it too much. I’m keeping things simple, automating the ‘big’ stuff but not trying to automate the micro details. And where other tools (such as Tasker/sharptools/AutoApps) are a better fit or easier to program, I don’t hesitate to use them along with, or even instead of, ST.

1 Like

I hear quite a few people say things like “oh my system has worked great for months,” and then I’ll ask “didn’t you get the emergency update in August? Weren’t you affected by the platform outages in July?”

And the answer is almost always “I wasn’t home when that happened” or “sure, but that was just a brief outage.”

My point is that there have been multiple issues that have affected all customers over the last six months. There can’t be anyone who had no outages during that time.

However, a lot of people just don’t pay attention if it’s something small. Or if they’re not home. Or if they get any kind of advance notice, even just a couple of hours.

In my case, I spend most of my life in one of two rooms, I need the home automation to work, and if it doesn’t work even for 15 minutes, I’m going to notice. An emergency platform update affects my ability to open my door or change the channels on my television or turn on a light. And it resets the maintenance free operating period.

I’m happy that people are happy with their systems and I’m glad they feel comfortable with them, but this is not yet anywhere close to a maintenance free system.

Just sayin’… :wink:


How do you expect the sensor/system to report status when the battery is dead. Perhaps this is not a good example :wink:

NO system goes without downtime. At work, we have systems on which lives depend that have downtime! We have mission-critical systems that have downtime! The email system that is the lifeblood of corporate communication goes down on a semi-weekly basis. That is in addition to the scheduled downtimes.

So I guess the real world has taught me not to expect relentless perfection. The emergency update? Yes I saw its effects for a moment. The outages? Yes I saw those too.

But the system was not out of commission for days in those instances. It was minutes, or an hour or so. And it came back up in good order.

I count that as robust and reliable, in general terms. Or if not perfectly reliable, then still far better than “soooo brittle”.

Actually, I think it’s a perfect example. ST should know the difference between “no status change reported” and “no communication at all.”

Right now it reports a dead/flat battery as a 100% full battery – that can’t be as the system was intended to work! Why won’t ST tell me when it can’t contact a particular node?

It apparently wakes up to report its battery percentage, though… So for weeks it was reporting ever-smaller battery percentages (say, 75%, 70%, 50%, 20%, 5%). Then, because ST didn’t hear from it for a few days, it assumed the battery was back at 100%? In my opinion, that’s a hub design flaw.

I don’t know the Zwave spec intimately, but evidently the door sensor is designed to contact the hub on some regular basis (to report battery percentage). When the hub hasn’t heard from the node in a while, it should tell me that.


Thanks, I didn’t know about SimpleDeviceViewer! Am checking it out now…

Right, you wouldn’t want an alert if the door was closed for weeks – but how about if the battery died? If the thing woke up once a day just to say it’s still alive, that would have a low impact on battery and a positive impact on system usefulness.

The sensor in question also reports temperature, so a day-long absence of readings should have been diagnostic for a dead battery.

Finally – the device list at showed the device as OFFLINE, so apparently the hub did know something about its health – it just wasn’t being reported by the app.

So I disagree that this is a Zwave limitation – it’s a poor design choice, and could be fixed.


What is brittle are the tiny plastic clips that hold the battery in-place in the presence sensor. After a couple battery changes, they give out and snap off. Then the sensor, with its loose battery, starts losing connectivity and opens doors in the middle of the night.

The solution is a small piece of open-cell foam. About 1/4" thick. Put it between the battery and the click-in plastic that closes the case.

1 Like